I think it is important to warm, at least the hands, for standardising the examination and for accuracy. A common example is examining for the carpal tunnel syndrome. Unless the hands are well-warm (important also for uisng the little finger as a benchmark), I ask the patient to get the hands - both, including wrists - under running (comfortably) hot water for at least a minute, or filling the sink and dipping the hands in to include at least an inch above the wrists. If no sink, I use a bowl. 

 

I believe some departments (Professor Erik Stålberg’s for example) have their own tables for correlating velocity with temperature. I do not, however, know if the curve or constant in the calculation applies equally to healthy as to diseased nerves.  Also, although we get an inkling on how much slowing is due to disease versus only to temperature from the shape of the response, I almost always eventually resort to proper warming - with water. I keep the trace before warming on the screen to compare with that after warming, to see the effect for the specific nerve first hand, to validate or - as often - negate my compensation estimate and also to intuitively add the observation to experience - useful when you just cannot do the warming.   

 

In the past at least, in Denmark, the arm or leg is placed under an infrared heater integral to the EMG machine, with monitoring of skin temperature, but I do not know if this is still used. 

 

Another method, from Finland, is getting the patient to don, while in the waiting area, thin plastic gloves up to mid forearm, then immersing the hands in a jar of molten wax. When the hands are taken out, wax solidifies in few mm-think layer on the gloves, warming the hands until the examination (expected in 10-20 minutes). Again, I do not know if this is still practiced - even in Finland.

 

Hot water is, I think, the most practical and effective and well worth the additional time. 

 

For the feet - especially if terminal latencies and distal velocities are crucial, I warm by immersing in a bucket (deplete industrial-25Kg-paint bucket, about 36cm diameter 36cm depth, I get from the estates unit of the hospital). Immersion includes about four inches above the malleoli. I would, however, be especially cautious in avoiding sudden dipping of cold-verging-on-cyanosis feet or hands; I would ascertain there is no contraindication to warming and raise the temperature of the water by gradually and carefully adding to the bucket. 

 

Although adding time, warming hands or feet is pleasantly perceived and appreciated by the patient and, to avoid guessing compensation for velocity, it is essential. But we will learn better from our colleagues.  

 

I think this question has impact beyond solely the diagnosis of demyelinating disease.  It's worth noting that the neurophysiological effect of cooling on extends beyond speed of conduction (latency & conduction velocity).  Among other effects, increased CMAP/SNAP amplitudes in a cool limb (due to prolongation of nerve depolarization) can impact interpretation, especially when combined with the latency prolongation effect.  Some quick examples of potential pitfalls:

• Axonal neuropathy could resemble a hereditary neuropathy with intermediate slowing, or could be read as normal.  
• Amplitude ratios (e.g., sural-radial) would be distorted, given physiologic tendency toward cooler distal limbs.  Side-to-side amplitude comparisons could have similar issues.
• Across follow-up studies, data would be difficult to compare (e.g., evaluating for evidence of axonal loss compared to a prior study performed 1 year prior; temperature status unknown).   

I think Dr. Ragi makes a very good point that we don't really know the impact of cooling on diseased nerves (e.g., what impact would/should cooling have on our interpretation of median-ulnar comparison studies in mild CTS?). 

Applying a correction factor for temperature (while sometimes needed) would not address potential differential effects of cooling on diseased nerves, nor would it address the known effect of cooling on SNAP/CMAP amplitude.  The distortion of electrophysiology is more complex than one of predictably prolonged latencies. These are some reasons- beyond the clinical situation of evaluation for demyelinating neuropathy- that support a best effort toward standard limb temperatures. 

Years ago we used infrared heaters, but we had safety concerns related to overheating. In recent years, we have used Hydrocollator heating pads. These are wrapped in a towel, are easily distributed along the course of most studied nerves, remain hot for most of the procedure, and are safe. Patients generally find them soothing. The tanks come in varying sizes. The pads last about a year before they need to be replaced. The downside is the initial expense, the need to periodically sanitize the tank, the space requirement for the tank (depends on the size), and the laundry from the towels. Most EMGers are probably familiar with these. 

Temperature has a big effect on NCS. In my fellowship we routinely saw people for second opinion regarding demyelinating neuropathies not responding to treatment only to repeat the EMG with warmed extremities and note the patient had a typical sensory-motor neuropathy not a demyelinating neuropathy. 


I have never had a patient complain of discomfort when warming ( I have used water as well as thermal packs from PT) . I have the patient warm while I am taking the history so it adds very little time to the study. 

Temperature has a big effect on NCS. In my fellowship we routinely saw people for second opinion regarding demyelinating neuropathies not responding to treatment only to repeat the EMG with warmed extremities and note the patient had a typical sensory-motor neuropathy not a demyelinating neuropathy. 


I have never had a patient complain of discomfort when warming ( I have used water as well as thermal packs from PT) . I have the patient warm while I am taking the history so it adds very little time to the study. 

If limb not warmed we will make too many diagnosis of carpal tunnel syndrome and too many unnecessary surgeries
beside other problems with cool limbs

The issue of temperature and clinical neurophysiology is certainly important and longstanding, but not without its controversy.  I am confident a number of my comments below will no doubt be considered controversial; not gonna be the first time.  I think it is important to appreciate that there are considerable nuances to temperature effects.   In the final analysis it is important for each practitioner to think about, and come to terms with the issues raised below as they are both very real and common place particularly for those of us in colder climates or during certain times of the year.
 
For the first 35 years of doing EDX I would meticulously measure temperature at the recording site as any other site made little sense to me. If the temperature was below 32 degrees C in the upper limb digits, or below 30 degrees C  at or about the ankle, I would warm the patient with carefully applied and monitored hot packs (patients with possible neuropathies will have altered sensation and MUST be monitored as they may not be able to tell you when their body part is getting dangerously hot-that’s why they are seeing you).  As previously noted in an above post, this process can be rather time consuming.   And, not infrequently, during the winter months no amount of prolonged heating may result in the desired temperature goal.  I became convinced by Charlie Bolton’s work that utilizing temperature correction factors was a “fools” errand”.  Specifically, he is the only person to my knowledge who performed well done and sufficiently large investigations to show that all of the previously derived temperature correction factors were done in healthy not pathologic nerves and hence of questionable relevance.  He not only documented different correction factors for pathologic nerves, but indeed, suggested different correction factors for different disease processes.  Hence utilizing correction factors in healthy individuals and applying them to potentially pathologic nerves (why else are people coming for an EDX assessment?) was a misguided practice at best.  Additionally, it must be recognized that utilization of any type of thermistor assessment, unless performed under the dermis, is recording the surface or skin temperature, and not the nerve itself.  There may or may not be a strong/direct correlation between the two, but one should never think that whatever reading you get on the skin is without doubt the temperature of the nerve itself (or muscle for that matter).  Is there any real data for this correlation or are we all making fundamental potentially erroneous assumptions?  There are a number of examples in clinical neurophysiology that are counterintuitive and the nerve vs surface temperature may be one of these issues.  Certainly, we have all measured skin temperature, found it to be quite “cold” yet the patient’s derived potentials are perfectly fine.  Alternatively, the skin’s surface temperature may be acceptable, however, induced SNAPs may be quite large and prolonged and refuse to change much despite warming efforts. 
 
After 35 years (apparently I’m a slow learner), I stopped measuring temperature and decided to take a step back and consider the “physiologic” aspects of the derived waveforms, i.e. what were their observed parameters in individual patients.  Further, could alternative approaches provide perhaps equivalent and sufficient data to arrive at the most appropriate diagnosis.  Additionally, an in-depth understanding of temperature effects (difference between generalized and focal cooling) as applied to a particular circumstance must be injected into the waveform’s interpretation.  For example, performing a so called “7&7” or median mid-palm SNAP may result in a mid-palm latency of 2.5 ms and a 14cm latency of 4.9 ms (trans-carpal time of 2.4 ms) with a wrist and mid-palm amplitude of 50 uV and 75 uV respectively.  Does this patient have a pathologic focal slowing of the median nerve across the wrist region? Well, maybe, and maybe not.  In the past, I would assiduously heat the patient up until the “cows came home” irrespective of the time required (way before the concept of RVUs and bean counter injected “productivity measures”).  Now, however, I make note of the findings: “large” amplitude waveforms, suggesting that there is most likely a temperature effect and hence the latencies may be potentially masking pathology across the wrist.  I am not aware of any peripheral neuropathies that “add” nerves, hence larger magnitude responses suggest some non-quantifiable temperature effects.  Specifically, the distal part of the nerve is more likely to be slowed and magnified more so than the more proximal and warmer neural segment resulting in an erroneous time difference across the wrist, or…. maybe not. So instead, let’s look at a different digit whereby the temperature differential may be less dramatic and perhaps more revealing of possible pathology.  Specifically, I like the first digit comparison or the ring finger comparison as well as the mid-palm mixed nerve studies.  On the other hand, what if a sural is 5.5 ms with an amplitude of 9 uV?  Is this solely a temperature effect or indicative of pathology?  To me, this is pathology and temperature.  Why?  Because, in my opinion, you can heat this nerve all day and no way are you going to get below 4 ms and the amplitude can only get smaller not bigger.  So heating will reduce the latency somewhat and reduce the amplitude, but I maintain the latency will still be long with a reduced magnitude response which equates to pathology, i.e. axonal loss and demyelination.  I believe the sural is more likely affected by generalized cooling (when adversely affected by temperature) as opposed to primarily focal cooling, while hand sensory nerves are subject to both types of cooling, but focal being the more prominent finding in the more distal segment about the MCP and IP joint regions.
 
I want to emphasize that I am not making any recommendations for what “should” or “should not” be done, and I expect to have no support for the above employed “physiologic” methodology- that’s ok.  But, how many practitioners continually measure temperature throughout the exam?  If not why not?  I expect a cold warmed extremity to cool rather quickly not only during the assessment of a single nerve but most likely between different nerve assessments.  Limb temperature is a dynamic not static process and how fast do you have to perform sequential stimulations prior to the area in question dropping in temperature to yield erroneous results.  In other words, how fast is the area you just heated cooling down and what is your time window to acquire results?  Is anyone employing continuous temperature monitoring in association with a feedback loop to modulate some type of radiant heating?  If so, that is truly a luxury most practitioners do not have and in particular, the gift of unlimited time per patient with no practical concern for productivity or equipment expense.  Or, are hot packs left on the limb under investigation throughout the assessment associated with moisture collection around the electrodes and along the path of stimulation possibly potentiating stimulus artifact or other adverse electrophysiologic/instrumentation consequences?  At what temporal point of not reaching your predetermined temperature goal (whatever that is) do you “call it” and punt?  Obviously, I am aware of what is “recommended” in the literature.  Further, I am also rather confident that most practitioners do not have laboratory certified, calibrated and maintained temperature measuring devices routinely standardized against known temperatures.  Also, what is the actually defined accuracy and standard deviations of most readily available and importantly affordable thermistors? We have three thermistors and they all record different temperatures from the same location.   Are we really measuring what we think we are?  Are we really being as precise, accurate, and quantitative as we think we are, or are there underlying confounding factors operational of which we may be only marginally aware?  However, if it is one thing I have learned in doing this stuff for quite some time, is that there are a number of ways to arrive at a correct electrophysiologic assessment ,and that we are clinicians first and our clinical acumen in association with objectively derived data must be interpreted together. I am simply trying to raise a number of concerns that each of us must come to terms with in the confines of our particular and practical time constraints as well as an appreciation that there is no single “approved” orthodoxy that must be followed to arrive at “the truth”.  It is important to formulate fundamental questions like those above to stimulate further discussion and hopefully some type of clarifying research with sufficient evidence demarcating a better defined pathway along which to proceed.   I think discussion like these are critical to delve into those nether regions where absolute certainty wanes and shades of gray predominate so that we each can arise through our own personal “phoenix” moment to arrive at a better understanding and hopefully individual clarity.
 

One easy, effective method for warming is microwavable "Medi-beads" packs. 2 packs are microwaved for 4 min and stay warm for about 30 minutes which is usually enough time for the NCS. They are large enough to wrap a hand or foot. I have my MA warm the patient for about 10 min before we start and we can usually maintain target temperatures which I monitor continuously. The Medi-beads are inexpensive and can be found on Amazon for about $25 each. Credit Rick Logigian and his team at URMC for the idea.

Thank you Dr. Dumitru (and others) for this thought-provoking post. Can you (and others) expand on your opinions/clinical experience on the effects of temperature (cold) with regards to comparison studies for median neuropathy at the wrist. Specifically, for example, if you are doing a radial/median comparison study in the setting of a cool or cooling limb – do we think the median and radial nerves are influenced in an equal matter thus giving a fair comparison? Or may there be focal cooling effecting one nerve more than the other? Or, as you stated, do we think that normal nerves are affected differently than pathologic nerves? Or, again as you stated, perhaps the time differential between stimulations is influenced by an actively cooling limb. I appreciate we are going into the world of theoretic plausibility and speculation. However, in your opinion, do you think we can accurately interrupt the comparison study without being confident the hand is warm throughout the study?  

Dr. Alessia, what a wonderful series of comments addressing both the explicit issues of currently employed clinical studies, and the implicit aspects of the “law of unintended consequences” regarding underlying pathology in the context of temperature related effects.  Please permit me a point of clarification regarding my post as it relates to the cautionary note pertaining to the utilization of temperature correction factors.  Specifically, as stated, this work was done by Charlie Bolton, M.D. with whom I had many pleasant conversations regarding his research, methodology employed, and ensuing recommendations.  He was primarily concerned with the empirically derived data from disease/non-disease nerves, and expressed concern over the use of correction factors which was very popular at the time; this practice has since fallen out of favor thanks in no small part to his work.  He did not, however, postulate as to the relationship between disease/non-disease sodium/potassium channel kinetics.  Although possible, it is perhaps unlikely that a disease process alters the ion channels’ protein composition and their response to temperature, but rather other issues such as impedance mismatches for varying degrees of demyelination, tissue conductances (intra-/extra-neural fluid), axonal factors, and perhaps other ill-defined tissue/nerve interactions operational in different disease states.  The point, therefore, is that cooling a normal nerve and calculating mathematical correction factors relating velocity and temperature do not take actual disease processes and a myriad of other factors into account, and hence should not be used because they do not yield the actually calculated correction factors in different disease processes substantiated empirically.  Further, though it is not actually known, that despite extraneous disease influences, it is likely that the particular kinetic responses of ion channels to temperature fluctuations remain similar between normal and different pathologies for the reason noted above.
 
I personally feel it is paramount for us to indeed embark on speculative ventures to push the boundaries of what is known, explore what isn’t known, and that which is assumed, i.e. asking fundamental questions to hopefully formulate excellent research investigations and find answers.  Pursuant to your excellent series of questions, permit me to categorically speak solely for myself and say: “I don’t know.”  Additionally, I am indeed making an assumption when utilizing comparison studies that the referential latency differences denoting pathology remain applicable.  The reason for these assumptions is several fold:  1) I have not read any studies documenting altered ion channel kinetics as they relate to temperature and disease despite attempting to find such information (if anyone is aware of this literature please provide those citations), 2) utilizing these studies in patients with symptoms seems to have correlated very well in the clinical setting over the years, 3) I have not documented despite trying, to define a temperature difference between the medial/lateral aspect of the digits in question when utilizing same digit comparisons I noted in my initial post, or studies to document the clinical/EDX effects of said anatomic/temperature differences.  Similarly, when warming a nerve, a number of assumptions are made as well: 1) the surface temperature is exactly what the nerve’s temperature is or close to it, and there is no time delay between surface temperature alteration to induced heating and the subcutaneous regions and the nerve’s response in reaching equilibrium, and do neural vs surface temperatures reach equilibrium simultaneously, or is the nerve still changing while the skin is not, in which case when should you begin measurements?, 2) normal and pathologic nerves respond in kind to warming over whatever time-frame is deemed appropriate, 3) all patients respond similarly to heating an assumed cold nerve, 4) continuous or intermittent heating to such a level required to actually effect a change in the subcutaneous tissues somehow represents a physiologic effect, i.e. replicating a non-heating limb and how ion channels/surrounding tissues react to propagating action potentials and in turn, how this relates to the reference population whereby that referential data was not obtained, i.e. with some type of external heat application: is the data obtained in non-heated limbs to establish reference values not significantly different from values in heated limbs,  5) external heat application has absolutely the same effect on disease/non-disease nerves, or maybe some disease nerves respond at faster rates because of different tissue impedances, 6) the thermistor used is calibrated, accurate, and never goes out of calibration, batteries are fully charged until they just quit, and a failing battery has no adverse effects on the instrument’s ability to record accurate temperature.  Personally, I am prepared to accept all the above assumptions until such time as there is evidenced based medicine providing a clear path to follow.  I fully support each of us defining for ourselves that which we have found to work, are comfortable with, and assured of doing what is in the best interest of our patients.
 

Although i had to commit a nigh insignificant amount of time and cognitive effort in doing so, I really appreciated both of your posts on this dr. dumitru. It certainly is inconvenient to challenge assumptions. Your points about spacial (surface vs neural temperature) and temporal (how long after heating) factors are worthy of note. 

Dr. Robinson often campaigns for the comparison studies for this as a supportive reason because the ring and thumb digits (somewhat) serve as temperature controls by using them. 

Superb comments from all, and I do not find Dr. Dumitru's comments controversial, but rather most educational.  I also agree with Dr. Malhotra's comments that comparison studies for CTS are quite resistant to the effects of temperature. Henry Lew and  I studied this about 20 years ago with no difference.  Also Mike Andary and I put my hand in freezing water for a long time and measured the CSI then and after warming - there was no difference.  So I think we are genreally safe for comparison studies to not worry about temperature.

And using our clincial acumen as mentioned, considering both amplitude and speed in the clinical context is so important. .  

But the infrared heaters can still be very useful - for heating up lunch, making toast, maybe grilling hotdogs, and even warming up one's feet when arriving into work from the deep cold of Canadian winters.  (But seriously I can still see using them for some polyneuropathy studies).

Love these discussions!

The benefit I receive from these posts assures me that my learning curve becomes steeper as the years pass by; (borrowing from Dr Johnson) triumph of experience over dogma.
 
I agree with all that has been said but, with apology for moving sideways from the original topic of temperature, may I be allowed discussion with focus on one concept - came twice in this thread, and numerous times in other.
 
“Axon loss”
We use this term to denote reduced nerve conduction amplitudes. But is this correct? It can be: pathology that has caused degeneration of axons. But is does not have to be, where the pathology causes axons to “switch off” - through disturbance in ion-channels and possibly numerous other factors – most yet unknown. Axons, however, are, in gross-anatomy and micro-anatomy, still intact. Examples are in variants of Guillain-Barré, toxins, anaesthesia.
 
‘Axon loss’ allows no ambiguity, to be understood as axons have actually been physically lost, and may thus wrongly convey dire prognosis. Conversely, if it is not physical loss, might this not beneficially place the onus on clinicians and scientists to consider how the pathology, reflected even if only partly in the ‘switching off’, could be reversed?
 
This has also been discussed in Complete Axon Loss, Ulnar NCS and other, and I have also had the pleasure of discussing the same with Dr Dumitru, on 17 September 2008 in Rhode Island. Dr Dumitru did kindly listen and accept my point in principle – but discussion was restrcited as he needed to attend a formal teaching commitment (and I apologise if my memory is incorrect).
 
A possible way out of the conundrum is (as Professor Mary Reilly, London, had suggested to me eight years ago) to use “axonal pattern”, which I use. We can also say ‘axonal’, but in either, the report’s recipient might mutter “What exactly do you mean?” Where we are not certain of the pathology, could we not come up with better terms?
 

Thank you Dr. Dumitru and Dr. Robinson for these thoughtful insights which have important clinical relevance. Also, thank you Dr. Dumitru for explaining out you’re reasoning. There was more than one factor you mentioned that I would not even of thought to consider (“unknown unknowns”). With that, you have helped me become more rigorous in my reasoning on this topic.
 

Yes, I agree that there may be "sick" motor axons that are viable but too sick to generate a CMAP.  McComas called these "silent synapses."  So I am not sure that the CMAP size totally reflects how many motor axons are preserved - sometimes it might underestimate.  

We have, as a field, generally been treating axons as either dead or alive - a dichotomous system.  But maybe they have sick days and just stop working for a while due to restricted axonplasmic flow or other causes.  

It is also important to just keep in mind that there is a temporally dynamic relationship between CMAP amplitude and pathology, particulary axonal loss and or conduction block.  Specifically, how long does it take for adversely affected axons to continue to respond to our external depolarizing input compared to its ability to physiologically (our voluntary control) conduct an action potential through a pathologic zone.  Also, how robust is the process of collateral sprouting in incomplete lesions over what time frame is required to once again re-establish muscle fiber activation and hence a restored contribution to the CMAP.  What percent of conduction block is involved and its own temporal parameters to "recovery"?  Perhaps these so called "sick" axons are experiencing a type of transient metabolic conduction block that take some ill-defined time to recover.  Maybe, these presumptive "blocked" axons result in not strictly "denervated" muscle fibers but nevertheless may result in collateral sprouting because the muscle fiber somehow "senses" it is no longer in physiologic (metabolic-axonal transport) continuity even though there is no actual physical disruption. This begs the question, then, can this type of conduction block result in Fibs/PSWs because of the issue of metabolic disruption of axonal transport.  And, on top of all this, when are we examining the patient in relationship to all of the above dynamic processes.  So, a very complex set of circumstances that need to be considered at least to some degree pertinent to this primarily hypothetical discussion with possibly very real consequences.  This is what I would have communicated to Dr. Ragi back in 2008 had more time been available for this discussion.

Thank you Dr Dumitru. Invaluable follow-up.

Colleagues, 
As some of us know, Dr Dumitru and Dr.Robinson and I have been "discussing" this for decades.  Finally my group decided to study to stop talking and actually study and publish this.  
The abstract is below.  
Dr. Dumitru's comments on this are so in depth and correct I cannot only sound stupid so I will not try to expound on them. He said somethng like: "Why dont you get off your butt and collect data".  So we did!!

After I froze Robinsons hand and checked his impressive carpal tunnel syndrome and Martin Gruber anastomoses he decided to move to Canada since he liked the cold hands so much. :-)
Bottom line: If you do the Robinson Index on the hand; temperature does not matter much.  I think Robinson and Dumitru predicted this. 

Sorry I was so late to get in this conversation. The cold weather has slowed down my response latency!!!
Thanks



1. PMID: 34532622

Spartan Med Res J

•  
•  
•  

. 2021 Aug 30;6(2):25941.
 doi: 10.51894/001c.25941. eCollection 2021.

Can Carpal Tunnel Syndrome be Appropriately Diagnosed in a Cold Hand?

Michael T Andary ¹, Drew B Parkhurst ¹, Maurice R Bernaiche ², Jose S Figueroa ³, Lata Kumaraswamy ⁴, Suzanne M Manzi ⁵, Ryan A O'Connor ¹, Ingrid P Parrington ⁶, Jim R Sylvain ¹
Affiliations expand

• PMID: 34532622
•  
• PMCID: PMC8405284
•  
• DOI: 10.51894/001c.25941

Free PMC article

Abstract

Introduction: The diagnosis of carpal tunnel syndrome (CTS) with nerve conduction studies traditionally involves warming the hand to avoid misleading prolongation of distal latency (DL). Comparing the median nerve DL to the ulnar and radial nerves using the combined sensory index (CSI) has been reported to improve the accuracy of CTS diagnosis. During this study, the authors examined the effect of hand temperature on the CSI and diagnosis of CTS.

Methods: The authors conducted a prospective, controlled, cohort study with 20 asymptomatic control patients and 21 symptomatic patients with confirmed CTS. Symptomatic patients underwent nerve conduction studies with the CSI calculated under both cold and warm conditions.

Results: Control subjects with warm hands had an average CSI of 0.0 milliseconds (ms), and -0.3ms with cold hands. CTS subjects with warm hands had an average CSI of 3.2ms, and 3.7ms with cold hands. Although hand temperature was shown to slow sample latencies, differences calculated with the CSI did not misclassify any of the 41 sample subjects.

Conclusions: During this study, cold temperature did not result in misclassification of either control patients or CTS patients when CSI was diagnostically used. Based on these results, peak latency comparisons in cold hands can be considered as diagnostically reliable as under standard hand temperature ranges for the diagnosis of CTS, with caution warranted in borderline cases. This diagnostic technique can save time for the patient, physician, and care team without compromising quality of care. Future larger sample blinded studies at multiple electrodiagnostic sites are indicated.

Keywords: carpal tunnel syndrome; electromyogram; nerve conduction study; temperature.

This is excellent. Thank you Dr. Andary!
I still remember how painfully cold my hand was when we froze it for a long time - but my CSI was pretty much the same cold vs. warm. 

I have been enjoying the learned discussion from Drs. Dumitru, Robinson and Andary on the topic of temperature effects on the nerve conduction and in the mean time learned a lot.

Dr. Dumitru, does the transient CB you described in "sick" axons explain the fib/PSWs we see in the UMN process. Could this be related to anterior horn cells not providing the "metabolic-axonal transport" continually despite having normal CMAP ampliltudes.I have never been able to understand or explain why there would be fibs/PSWs in patients with UMN syndromes.

In all the discussions about temperature effect CV and amplitude are discussed. However, there is no discussion of the duration of the CMAP. This could be severely affected as well. A study in Muscle and Nerve about CTS several years ago by Cleland et all (2003) discussed about the dispersion of the median CMAP and sugested that duration +/> 9ms is never seen in even the most severe CTS and is a hallmark of CIDP with a specificity of 94% and sensitivity of 78%. This increase in temporal dispersion would be more important to consider for making a diagnosis of generalized vs focal disease process that has relevance to treatment. 
 

I want to be a bit careful here in that we are now discussing two rather different topics:  temperature, and some type of poorly substantiated hypothetical conduction block.  I would defer to the moderator(s) regarding any type of “split” between topics and discussion streams.
 
In my opinion, SNAPs are affected much more dramatically by temperature than CMAPs.  There are several physiologic reasons for this difference in response to temperature reductions.  Obviously, muscle tissue is both more proximal and vascular compared to digital sensory nerves.  The dynamic nature of muscle’s vascular tissue to temperature alterations, and the warmer blood flow helps to modulate significant temperature swings.  Further, most muscle tissue is surrounded by other muscles all helping to maintain a more even temperature distribution compared to digital nerves more-so exposed to, and subsequently more influenced by temperature reductions (nerves close to the surface with limited vasculature and subcutaneous tissues).  Further, there is a wider distribution of NCV velocities (slowest compared to fastest) for nerves compared to muscle conduction velocities, thereby mitigating significant temporal dispersion of individual fibers for muscle with comparatively less temporal dispersion and less polyphasicity.  This can be seen normally with rather dramatic reductions in SNAP amplitudes and increases in waveform duration with more proximal stimulations compared to CMAPs, which change way less in both amplitude and duration.  What I have noticed with dramatic reductions in purposeful (artificial) hand cooling is that the CMAPs hold their own with not that much alteration in configuration until they just stop conducting.  The velocity difference in nerve compared to muscle along with the manner in which faster compared to slower conducting nerves respond to temperature reductions along with the slower closing of sodium channels all summate to yield larger distally recorded SNAPs with longer durations, i.e. focal cooling effects with a decrease in velocity dispersion but an increase in slowing leading to less phase cancellation and amplitude increases.  There is a slightly different reaction to a more longitudinal expanse of neural cooling (less dramatic amplitude increases but perhaps more longer durations and more dramatic latency slowing).  But, you do not typically see an increase in polyphasicity again suggesting an overall slower, but a more uniform velocity distribution.  An increase in response polyphasicity usually suggests a differential effect on velocity distributions with an ensuing increase in temporal dispersion and commensurate reduction in amplitude, but relative preservation of waveform area irrespective of the underlying disease process, i.e. CIDP, Diabetes, etc.  This is one distinguishing characteristic of conduction block and demyelination, i.e. a reduction in area for the former but preservation of area for the latter.  Caution should be exercised in measuring CMAP areas in various pathologies compared to “normal” and extrapolating that information to supposed temperature effects, particularly in muscle tissue which is less well characterized than for nerve.  Again, as I noted in my initial post, this is a very complex issue with a multitude of nuances and various areas of controversy and a paucity of actual basic science information.
 
At the risk of becoming a fool and treading where angels fear to go, speculation on “upper motor neuron” Fibs/PSWs has always been of interest to me.  I am a firm believer in Fibs/PSWs never being “normal” and reflect that the single muscle fiber resting membrane potential has become unstable secondary to the generation of non-wild type sodium and potassium channels with distinctly different kinetics than wild type sodium/potassium channels (Thessleff did significant work on this topic with wonderful basic science sodium/potassium channel labeling demonstrating all of this with intracellular recordings of IAPs).  Unfortunately, the majority of clinicians doing clinical neurophysiologic are not aware of this foundational work.  Hence the mistaken idea that somehow and sometimes the detection of Fibs/PSWs can be a “normal” finding.  There is a difference between “subclinical” and “disease free”.  The modulation of these various ion channel production by the muscle cell is directed by some ill-defined substance mediated by slow axonal transport, hence the time delay between cessation of slow axoplasmic flow, length of preserved axon and the substance contained therein, and the replacement of wild type ion channels all result in various time delays before documentation of first increased insertional activity to be followed by sustained Fibs/PSWs.  Reinnervation has the reverse effect of re-established flow, replacement of non-wild ion channels and reduction in Fibs/PSWs followed by electrical silence at rest.  So, it may well be that there is some substance in “higher-order” neurons directing in some modulatory manner the lower order neurons with ensuing effects on the final end-organ, e.g. the single muscle fiber.  Again, this is way too complex to explore in anything other than very superficial means in online posts.  For those that are interested in this topic, the majority of this basic science material is covered in a much more in-depth manner in either of the editions of my book on Electrodiagnostic Medicine.  I do apologize for these long posts.  Unfortunately, the complexity of the material requires more than cursory responses out of respect for those interested in the material.
 

Again an excellent and very thoughtful discussion, which is illuminating to me.

I know this off topic for temperature,  but I wish to ask one other question about fibs/psws ever being "normal" or not.

It's been well demonstrated that MUAP amplitudes and durations become larger with age.  This means that there has been some loss of motor neurons as we age and subsequent reinnervation.  Could we sometimes be lucky or unlucky enough (depending upon perspective) to do needle EMG soon after the normal loss of a motor neuron and record from it's recently denervated muscle fibers, before reinnervation?


 

Dr. Robinson brings up an interesting point that perhaps falls into the category of “philosophy” of disease states.  Should one consider the aging process as a subclinical (or in my case not so subclinical) “disease state”?  I would purposefully side-step this question and consider the point from the perspective of “cellular degradation” irrespective of cause.  From my perspective, if an anterior horn cell is lost secondary to motor neuron disease or aging, the anterior horn cell has been compromised from a cellular perspective.  As a result, the ensuing muscle fibers will at some point begin producing non-wild type ion channels, develop an unstable resting membrane potential and generate spontaneous electrical discharges we refer to as Fibs/PSWs.  However, if the process of collateral sprouting is robust enough, we may never see them (Fibs/PSWs).  But, as Dr. Robinson suggests, we may indeed record these waveforms prior to successful reinnervation.  Absolutely.
 
Clearly, there is a distinction between the issue of clinical, concern or not, with respect to the detection of Fibs/PSWs, and whether these waveforms can be considered “normal” or “abnormal”.  As noted in a previous post by me, I consider the detection of “real” Fibs/PSWs (not mistakenly identified end-plate spikes that “look like” Fibs/PSWs) to be indicative of an unstable resting membrane potential.  An unstable resting membrane potential is not what would typically be considered as a “normal” condition, ok at least not “normal” by me.  Also, one of the generally accepted hallmarks of the needle EMG is a quiet baseline at rest aside from end-plate activity.  The corollary to this principle is that disease free muscle tissue should have a stable resting membrane potential and not display spontaneous discharges at rest on needle EMG.  This concept has certainly been a well-accepted axiom in needle EMG since before my time.  Of course, dogma does not equate to fact, and all axioms should be challenged from time to time.
 
Sooooooo… what if you detect one or two Fibs/PSWs in a single muscle. Well, clinical context is everything.  Let me ask a few questions not directed at anyone in particular;  What would you think if you saw these waveforms in a 95 year-old vs a 45 year-old?  What would you say in your report?  What would you do, if anything?  Does the examined muscle make a difference, i.e. foot intrinsic/paraspinal vs biceps brachii?
 
Let me personally thank my esteemed colleague from north of the border for continuing to challenge my “little gray (grey for my Canadian friends) cells”, help maintain my neural plasticity, and in general make me smile.🤠
 

Continued excellent discussion from my colleague with the very large number of grey cells (grey is an excellent colour, BTW).

I guess this is, at least in part, a discussion of terminology. 
 
Normal comes from the Latin meaning right-angled.  Things that are correct and right in the world. 

Healthy is somewhat different - it refers to being sound of body and being able to discharge all functions efficiently.  So I may be healthy, but have never been accused of being normal.

So perhaps, occasional fibrillations can be seen in muscles of healthy people, but we don't consider them "normal".  That's a distinction I could live with.  And it reinforces the need to take everything in clinical context.



 

The area under the curve you made is an excellant point in differentiating the CIDP vs. diabetic neuropathy or CTS. Thank you for this observation and is extremely helpful in clincial practice.

The point you made about the muscles being surrounded by other muscles and hence not affected by temperature is a bit confusing as the muscles we record either in the hand or foot are just under the skin and are not surrounded by any other muscles. I do see increase in duration with the cool limb with increased amplitude of CMAP with increased latencies when temperature is in the mid 20s as opposed to low 30s (especially being in northeast). This may be less of an issue in Texas. Definitely the area under the curve does not change with proximal stimulation in these nerves. 

I agree with you completely that presence of abnormal SPA is never normal except as you note possibly in the foot muscles (PSW but not Fibs), As you note clinial cotext is extremely important. Occasionally I may see a patient who came in for evaluation of radiculopathy but has few PSW in a single muscle without any pattern. I point out that this is there but I am not sure I can say that this is due to a root compression and may be due to local factors such as local trauma etc. This may be the example Dr. Robinson talked about not being normal but healthy:)

Thank uou for these posts that are very helpful in understanding of the basic neurophysiology.
 

https://i.imgur.com/OFYg4nT.jpg

As I stated in a previous post on this topic, the entire issue of temperature is both complex and subtle.  Please permit me a few comments in reference to one of my posts to ensure that there is no misinterpretation.  I did not state, “…muscles being surrounded by other muscles and hence not affected by temperature…”, but did state, “In my opinion, SNAPs are affected much more dramatically than CMAPs.”  Clearly, my original statement is saying muscles/CMAPs are affected by temperature reductions, but not to the same extent as SNAPs.  Just one of the multiple reasons I gave for this relative resistance is because of the comparatively more proximal nature of where CMAPs are derived, i.e. in the hand: APB.  Specifically, the APB itself is rich with a blood supply and “surrounded” by the OP, FPB (superficial/deep heads), and lying over the AP and more deeply FDI, all of which also contain vascular tissue and in part comprised of circulating blood maintaining some degree of a heat exchange mechanism replacing cooler with warmer blood.  This heat exchange mechanism keeps the thenar eminence comparatively more warm than the proper digital nerves and helps to modulate temperature fluctuations to a greater degree than the subcutaneous relatively less vascular digital nerves which are “surrounded” by tendon and bone.  That is not to say there is absolutely no temperature effect for the APB’s CMAP as it is just subcutaneous on its superficial aspect, but is “surrounded” (but not enveloped) by a thermal heat exchanger.  Therefore, from a three-dimensional perspective regarding anatomy, the same may be said to varying degrees about the ADM, AH, but much less so about the EDB (for this and other reasons) which is why I hardly ever study the EDB muscle, but that is a topic for another discussion.
 
 
Consider the examples of my APB (see Figure) recorded with a temperature approximating 34 degrees C (upper trace) and again at 19 degrees C (bottom trace) following a rather uncomfortable equilibrating ice bath:  purposefully pushing the limits to demonstrate the undeniable CMAP consequences of icing and an attempt to overwhelm some degree of thermoregulation.  You may, but are unlikely to encounter such a dramatic temperature reduction in clinical practice. The recording and stimulating electrodes/parameters were not altered between the two circumstances.  This is the extreme effect of temperature on CMAPs I noted above in one of my previous posts with associated rationales.  Specifically, there is a prolongation in DML (4.1 ms to 5.4 ms) with an increase in CMAP negative spike duration (4.5 ms to 6.0 ms), essentially no change in negative spike area (21.3 ms-mV to 22.8 ms-mV) but unlike focal neural cooling, a reduction in base-to-peak amplitude (8.2 mV to 5.9 mV).  By the time environmental conditions are such that the muscle tissue is rendered “cold enough” ( this is cutaneous temperature-who knows what the intra-muscular temperature actually is), so is the proximate peripheral nerve and intramuscular nerves, NMJ, as well as, the muscle tissues’ action potentials (this last parameter does not contribute to the DML).  The end result is a significant slowing of peripheral/intramuscular neural conduction, NMJ transmission, and muscle fiber conduction velocity.  Therefore, the CMAP’s DML is delayed (neural conduction/NMJ,) with less temporal synchrony of action potential arrival times as well muscle fiber conduction slowing all resulting in an increase in CMAP duration and a reduction in CMAP amplitude, not an increase.  The CMAP “area under the curve” however, remains the same implying no significant phase cancellation or conduction block (this was addressed by me in an earlier post), but rather an increase in temporal dispersion (totally different mechanism than SNAPs recorded at progressively more proximal locations yielding both amplitude and area reductions; e.g. ulnar SNAPs).  The nerve/muscle effects on my CMAP is in effect the physiologic effect approximating “general” cooling as opposed to “focal” cooling. I want to emphasize that this dramatic effect of temperature on CMAPs occurs when the cooling is so significant as to challenge the thermoregulatory effects of the “surrounding” muscle tissue (circumferential and underlying tissues; prolonged ice bath to achieve this effect, i.e. 14-15 degree reduction in surface temperature) and again, perhaps not as dramatic as one is likely to encounter clinically.  I do want to point out that if one only has the cold CMAP (Figure) in isolation, one may suspect qualitatively but cannot categorically state quantitatively that either the amplitude or duration are without question adversely affected for the waveform in the Figure: i.e. if you ignored the DML which is a dead giveaway for a PN, Temperature effect, or both.  This is because we have such wide standard deviations for both of these CMAP parameters without well-defined upper limits particularly regarding CMAP duration.  If you encounter someone with an APB surface temperature of 19 degrees centigrade like mine, then absolutely warm them up if for no other reason than compassion as I can tell you this is not just uncomfortable but painful.
 
Finally, thermoregulatory effects are also dependent upon one’s ability to acclimate to the environment.  Individuals in warmer climates may not be as capable of thermal compensation compared to counterparts in more cold climates and display potentially significant effects at temperatures of no consequences to others more used to the cold.  Also, those persons in warmer climates experiencing unusual cold may not have appropriate clothing/gloves and present with more dramatic temperature effects than those persons more knowledgeable and experienced/prepared in dealing with the cold (e.g San Antonio vs Philadelphia).  Fear induced peripheral vasoconstriction can also play a role in limb temperature.  Hence, various individuals may have varied temperature effects irrespective of their geographic locations and time of year.
 
Finally, to return to the issue of Fibs/PSWs, I make a distinction pursuant to the tissue in question between “healthy” (absence of pathology), “subclinical” (pathology present but not clinically manifest), and “clinically manifest” (pathology present and manifesting through History and or Physical Examination/laboratory data).  As noted previously and in my opinion, Fibs/PSWs (Fibs=PSWs) is objective documentation of an unstable single muscle fiber resting membrane potential which I consider to be neither “normal” nor “healthy” for that single muscle fiber.  That is not to say that a totally “healthy” person otherwise, cannot have a few Fibs/PSWs in a muscle that has had some “minor” one-time local trauma adversely affecting a few muscle fibers or axons and your needle just happens to be in that location prior to collateral sprouting, but after the manifestations of Fibs/PSWs.  On the other hand, how do you know that is the case?  How do you distinguish this circumstance from someone who in fact does have a slowly progressive “subclinical” pathologic process adversely affecting a few muscle fibers/axons at the time of your examination, but otherwise “appears healthy”?  In fact you don’t “know” which possibility is operational beyond any doubt.  As a previous poster stated something to the effect of:  make note of it, document the finding, comment on it within the clinical context, and suggest a prudent follow-up.  My take on Fibs/PSWs in foot intrinsic/paraspinal muscles has been well delineated (Dumitru D, et al:  Prevalence of denervation in foot intrinsic and paraspinal musculature.  Am J Phys Med Rehabil 2001:80:482-490). 
 
As always, these discussions are quite illuminating but can be frustrating for both writer and reader, particularly when attempting to address the nuances of complex neurophysiology that have not been thoroughly studied and hence anecdotal data must suffice.  Further, attempting to discuss these complex and nuanced topics with short pithy sentences is also difficult.  As previously stated, these are solely my opinions, and I am not attempting to promulgate some type of dogma which in the final analysis is no substitute for actual data: Dogma is not equal to Fact.  It is my hope, however, that discussions such as these serve as a stimulus to raise further questions, formulate research ideas, and eventually lead to clarifying data.  My apologies for any lack of clarity and/or confusion in my previous discussions.
 

Although some of this is way over my head, as a technologist, I always find it fascinating reading the "greats" in EMG.   The topic of temperature is very enlighting.  I always LOVE reading Dumitru, Robinson and Andary :)  

Thank you so much for sharing :)   

Carrie Ford

Thank you Dr. Dumitru for the illuminating explanation fo the effects of temperature on the different parameters and fibers in NCS. It is extremely helpful to have your learned explanations to understand basic neurophysiology. 

Colleagues....

Once again the Robinson Index (CSI) proves invaluable. Cold or warm hands.

For purposes of EMG Lab accreditation, one needs to include skin temp dorsum of hand > 32 degrees C. And dosum of foot > 30 degrees C. Use a plain old infrared thermometer and get your temp in a millisecond by pointing to the skin. 

Look, there isn't a perfect system to warm the skin, and yes  skin temp does not equal nerve temp. 

I've tried every way and I've settled into a warm exam ROOM and a hair dryer. Yep, a fancy machine called a hair dryer.

BUT, I assure you that any warming method you use  can cause first degree burns, especially in people with neuropathy in feet, so check temp against your skin constantly the way you would check the water temp before submerging a baby in a bath. 

And don't forget that temp is just as important in
RNS of ulnar to ADM, (cooler temp reduces the decrement on RNS).

OR CAMP  amplitude in LEM is affected by temp.
 (cooler skin temp  means raising or even normalizing the low initial CMAP amplitude in LEM.

The bottom line is there's no perfect method of warming, but I'm 100% on the side of warming.

Years ago I shared a vignette with Dr. Robinson about a physician who clinically in his early 40s and had developed a sensory neuronopathy after receiving cisplatin for testicular CA.

However he had an outside NCV study that showed
"normal" data but the NCVs were tested when his skin 
was cool, so the SNAP amp in sural was about 7uV
and CV was about 42m/s.

When I did the study and I warmed the patient's foot, the sural SNAP amp was around 3uV and CV was around 49m/s.  So in his case, keeping the skin cool actually normalized the data; thus, no evidence of a sensory
neuronopathy was described on outside EMG/NCV report.
As I said clinically and symptomaticaaly he had an obvious sensory neuronapathy by history and physical, and by warming
his skin I was able to corroborate the sensory neuropathy by NCV's.

And remember, if you submit an EMG report for grading for EMG lab accreditation and no temperature is documented, we will send the study back to you for you to remediate it and ask you to report a temperature.

Great discussions by everyone. I doubt there is much that
generateds more discussion than temperature.

It is the bane of existence of people who do NCVs. 

My rules of the road....

Remember to keep the ambient room temp warm.
While it might not be kosher, a small electric heater does
wonders. 
And then your trusty little hair dryer to get skin warmer if you need to,
and infrared thermometer! 

That's my two cents worth!!! 

 

 

We have talked at length about the ambient temperature and focal temperature related to focal cooling. However, the disease processes such as Raynaud's or severe arterial disease cool the whole limb temperaure including the muscle and the nerve. Be aware and sometimes you may not be able to warm these patients and have the abnormal parameters of the NCS.

As Dr. Zimnowodski pointed out about the heating of a limb that has disease, it is very easy to burn the patient to aquire appropriate temperature. When you are warming a limb it is always important to feel the temperaure of the water, have your hand on the area you ar trying to warm with the "hair dryer" to make sure even in healthy poeple that you are not burning their skin. 

Other point that needs to be made is about the place of recording of the temperature and place of recording. If the temperature probe is on the wrist or dorsum and you are recording from the digit thgat is significantly colder, the data may not be appropriate. However, when the data published is analyzed, there is no mention of the point of temperature recording in most of the studies. That leaves the question Dr. Dumitru raised about temperaure recording in the beginning of this discussion as not as relevant. My two cents.